SBELTS: A model of soybean production under tree shelter

Farmers in the North Central region of the United States often are reluctan t to use shelterbelts because of inadequate information clearly showing the ir benefits. We developed a computer model, called SBELTS for ShelterBELT a nd Soybeans, that simulates the influence of a shelterbelt on soybean (Glyc ine max L.) production across an agricultural field in the midwestern Unite d States. Objectives of this study were to 1) describe the structure of SBE LTS, 2) present model simulations, and 3) discuss model limitations. SBELTS is composed of three submodels. The first submodel produces characteristic s of a shelterbelt that are passed to the second submodel that estimates da ily windrun at user-specified distances leeward (away from the prevailing w ind) of the shelterbelt. Estimated daily windruns are merged with other mic roclimatic information to produce weather files for each specified distance . Finally, the third submodel uses a soybean growth/yield model to estimate soybean yield at specified distances, and the yields are averaged. Data co llection was not a part of the present project, so SBELTS was evaluated by comparing predicted results with published information. SBELTS was used to predict soybean yield across a field leeward of a 7.6 m tall shelterbelt, a nd the predicted yield curve compared well with published yield curves. The sensitivity of SBELTS to variation in rainfall was evaluated by predicting yields for 3.8 m tall and 7.6 m tall shelterbelts in wet, normal, and dry years. Results showed no shelterbelt influence in wet years, some influence in normal years, and a sizable influence in dry years. Results showed that the 7.6 m shelterbelt had more influence than the 3.8 m shelterbelt. Altho ugh SBELTS has limited use, it is the first step in the development of more advanced models that will be able to simulate production of soybeans and o ther crops under the influence of shelterbelts on a variety of soil types.